It should be recognized that writing has a great many limitations. That which is written is lacking a great deal of the context that is present with face-to-face communication, and so it must be interpreted by the reader. This often goes awry. The lack of tone and body language that written language comes with makes it hard for people to read irony, humor, and sarcasm in written texts, since it can be a challenge to encode enough information for such communication to be picked up by the reader. Often the challenge and time-consuming (and somewhat expensive) nature of writing can mean that a writer often chooses to omit writing about what is obvious to both the sender and the recipient, not recognizing that what is obvious to them may not be obvious to a reader from outside of their shared context and understanding. (This can, it must be admitted, sometimes be a feature and not a problem if the writer and recipient are depending on having a private context to mutually understand what an outsider would not recognize.)
There are even more exotic limitations when it comes to writing. The comprehensibility of writing depends on a reader knowing both the proper way to pronounce whatever is being written in whatever alphabet, ablaut, syllabary, cuneiform, or pictograph system, or any other system that could be devised, and also to understand what is being meant. It is by no means obvious that what is written down will be understood by a later reader. There are many writings that we possess that we cannot understand because we cannot read the writing system used by the author, though decipherment is such a powerful goal that a great deal of effort continues to be spent on trying to understand these texts by relating them to languages we happen to know, a task that can bear fruit when the right target language is shown. Even when we can pronounce a language, like we can pronounce Etruscan because it is written in an alphabet we can understand (being based on the Greek alphabets), we may not be able to understand it because it is not sufficiently like any language that remains for us to be able to translate it, nor may enough writing survive for us to attempt to get at the language through brute force methods of statistical analysis.
Yet despite these limitations, a great deal of writing goes on, and it is well worth understanding why. Writing is meant to fix texts in place. This seems somewhat obvious to us, but it is not always so. A great deal of ephemeral communication exists in written words that are spoken into the air, heard from that same air, and acted on (or not), but which do not demand to be fixed in place nor do they have any lasting worth. But the desire to keep track of and audit and control behavior leads quite naturally and inevitably to fixed writings. Friends can easily make verbal agreements that they will hold themselves to that commit them to going to certain places and doing certain activities at a certain time. No further record is needed for it. But if we are going to collect taxes from strangers or verify that the right procedures are followed in a particular sales transaction, we are going to need some sort of record that can be consulted later on, which requires communication to be fixed in some fashion. For all of its limitations, writing is a particularly handy way for communication to be fixed for future reference.
Let us compare, for example, a written contract or transcript with a recorded sound file, as far as providing fixed communications. The written contract or transcript has many advantages over the recorded file. It requires no specialized hardware to listen to, far less space to store either in its original or copies, and can be consulted far faster as we usually read far faster than we can listen to something. The greater technical difficulty of recording sound and video has meant that we have, predictably, a far smaller and far more recent collection of such recordings than we have of writing, which goes back thousands of years. And even the current large amount of recorded material we have is far more vulnerable to be lost than writing is. And yet we record and write information, hoping that it is not written in the wind, but that we have some fixed record of our lives and thinking for others separated by time and space to recognize and hopefully appreciate.